This dataset consists of word profiles for George Washington's
manuscripts. This dataset is the FiftyWords dataset, remapped to 25
classes. The data was flipped left-right so that it would not be
recognized.
